Title:
FREE FORM INPUT FIELD SUPPORT FOR AUTOMATED VOICE ENABLEMENT OF A WEB PAGE
Kind Code:
A1


Abstract:
Embodiments of the present invention provide a method and computer program product for the automated voice enablement of a Web page with free form input field support. In an embodiment of the invention, a method for voice enabling a Web page with free form input field support can be provided. The method can include receiving speech input for an input field in a Web page, parsing a core attribute for the input field and identifying an external statistical language model (SLM) referenced by the core attribute of the input field, posting the received speech input and the SLM to an automatic speech recognition (ASR) engine, and inserting a textual equivalent to the speech input provided by the ASR engine in conjunction with the SLM into the input field.



Inventors:
Moore, Victor S. (Lake City, FL, US)
Nusbickel, Wendi L. (Boca Raton, FL, US)
Application Number:
12/099041
Publication Date:
10/08/2009
Filing Date:
04/07/2008
Assignee:
INTERNATIONAL BUSINESS MACHINES CORPORATION (Armonk, NY, US)
Primary Class:
International Classes:
G10L13/08
View Patent Images:



Other References:
Fielding, Roy Thomas. Architectural Styles and the Design of Network-based Software Architectures. Doctoral dissertation, University of California, Irvine, 2000
M. Hu et al., "Speech-enabled web services for mobile devices," in Proc. the International Conference on Semantic Web and Web Services, 2006
Voice Extensible Markup Language (VoiceXML) 2.1 W3C Recommendation 19 June 2007
Voice Extensible Markup Language (VoiceXML) Version 2.0 W3C Recommendation 16 March 2004
Primary Examiner:
BAKER, MATTHEW H
Attorney, Agent or Firm:
Nuance Communications, Inc. (c/o Wolf, Greenfield & Sacks, P.C. 600 Atlantic Avenue, Boston, MA, 02210-2206, US)
Claims:
We claim:

1. A method for voice enabling a Web page with free form input field support comprising: receiving speech input for an input field in a Web page; parsing a core attribute for the input field and identifying an external statistical language model (SLM) referenced by the core attribute of the input field; posting the received speech input and the SLM to an automatic speech recognition (ASR) engine; and, inserting a textual equivalent to the speech input provided by the ASR engine in conjunction with the SLM into the input field.

2. The method of claim 1, further comprising: receiving additional speech input for a different input field in the Web page; parsing a core attribute for the different input field and identifying permitted terms set forth in the core attribute of the different input field; generating a speech grammar for the different input field based upon the terms in the core attribute of the different input field; posting the additional speech input and the generated speech grammar to an automatic speech recognition (ASR) engine; and, inserting a textual equivalent to the additional speech input provided by the ASR engine in conjunction with the speech grammar into the different input field.

3. The method of claim 2, wherein generating a speech grammar for the different input field based upon permitted terms in a core attribute of the different input field, comprises generating a speech grammar for the different input field based upon permitted terms in a hidden title attribute of the different input field.

4. The method of claim 2, wherein generating a speech grammar for the different input field based upon permitted terms in a core attribute of the different input field, comprises generating a speech grammar for the different input field based upon permitted terms for a corresponding grammar listed in a core attribute of the different input field and also a prefix for the permitted terms.

5. The method of claim 2, wherein generating a speech grammar for the different input field based upon permitted terms in a core attribute of the different input field, comprises generating a speech grammar for the different input field based upon permitted terms for a corresponding grammar listed in a core attribute of the different input field and also a semantic indicator for the permitted terms.

6. The method of claim 2, wherein posting the additional speech input and the grammar to an automatic speech recognition (ASR) engine, comprises posting the additional speech input and the grammar to a communicatively coupled remote representational state transfer (REST) compliant ASR engine.

7. A computer program product comprising a computer usable medium embodying computer usable program code for voice enabling a Web page with free form input field support, the computer program product comprising: computer usable program code for receiving speech input for an input field in a Web page; computer usable program code for parsing a core attribute for the input field and identifying an external statistical language model (SLM) referenced by the core attribute of the input field; computer usable program code for posting the received speech input and the SLM to an automatic speech recognition (ASR) engine; and, computer usable program code for inserting a textual equivalent to the speech input provided by the ASR engine in conjunction with the SLM into the input field.

8. The computer program product of claim 7, further comprising: computer usable program code for receiving additional speech input for a different input field in the Web page; computer usable program code for parsing a core attribute for the different input field and identifying permitted terms set forth in the core attribute of the different input field; computer usable program code for generating a speech grammar for the different input field based upon the permitted terms in the core attribute of the different input field; computer usable program code for posting the additional speech input and the generated speech grammar to an automatic speech recognition (ASR) engine; and, computer usable program code for inserting a textual equivalent to the additional speech input provided by the ASR engine in conjunction with the speech grammar into the different input field.

9. The computer program product of claim 8, wherein the computer usable program code for generating a speech grammar for the different input field based upon permitted terms in a core attribute of the different input field, comprises computer usable program code for generating a speech grammar for the different input field based upon permitted terms in a hidden title attribute of the different input field.

10. The computer program product of claim 8, wherein the computer usable program code for generating a speech grammar for the different input field based upon permitted terms in a core attribute of the different input field, comprises computer usable program code for generating a speech grammar for the different input field based upon permitted terms for a corresponding grammar listed in a core attribute of the different input field and also a prefix for the permitted terms.

11. The computer program product of claim 8, wherein the computer usable program code for generating a speech grammar for the different input field based upon permitted terms in a core attribute of the different input field, comprises computer usable program code for generating a speech grammar for the different input field based upon permitted terms for a corresponding grammar listed in a core attribute of the different input field and also a semantic indicator for the permitted terms.

12. The computer program product of claim 8, wherein the computer usable program code for posting the additional speech input and the grammar to an automatic speech recognition (ASR) engine, comprises computer usable program code for posting the additional speech input and the grammar to a communicatively coupled remote representational state transfer (REST) compliant ASR engine.

Description:

REFERENCE TO CO-PENDING APPLICATIONS FOR PATENT

The present application is related to the following co-assigned U.S. patent applications, which are expressly incorporated by reference herein:

U.S. application Ser. No. ______, entitled “AUTOMATED VOICE ENABLEMENT OF A WEB PAGE” (docket no. BOC920070071US1 (83U)), filed on Mar. ______, 2008.

U.S. application Ser. No. ______, entitled “PROACTIVE COMPLETION OF INPUT FIELDS FOR AUTOMATED VOICE ENABLEMENT OF A WEB PAGE” (docket no. BOC920070072US1 (84U)), filed on Mar. ______, 2008.

BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention relates to the field of speech recognition and more particularly to voice input processing in a Web page.

2. Description of the Related Art

Speech recognition systems perform a critical role in commerce by providing an essential reduction in operating costs in terms of avoiding the use of expensive human capital in processing human speech. Generally, speech recognition systems include speech recognition and text-to-speech processing capabilities coupled to a script defining a conversational flow. Consequently, speech recognition systems can be utilized to provide a voice interactive experience for speakers just as if a live human had engaged in a person-to-person conversation.

Speech recognition systems have proven particularly useful in adapting computer based information systems to the audible world of voice processing. In particular, while Web based information systems have been particularly effective in collecting and processing information from end users through the completion of fields in an on-line form, the same also can be said of speech recognition systems. In particular, Voice XML and equivalent technologies have provided a foundation upon which Web forms have been adapted to voice. Consequently, speech recognition systems have been configured to undertake complex data processing through forms based input just as would be the case through a conventional Web interface.

Integrating speech processing with Web pages of a Web application is not for the feint of heart. Expertise is required not only in the field of Web page and Web application development, but also in the field of speech application development. Integral to speech application development is the design and use of a speech grammar reflecting the intended use of the speech application. In that expected information for different fields in a Web form can vary from field to field, in many instances, a different grammar must be generated for each field.

Thus, the speech development component of a voice enabled application often occurs concurrently with the development of the Web application and requires coordination between the speech developer and the Web developer. Only then can both be combined into a single voice enabled Web application. Given the complexity of this task, voice enabled applications elude widespread adoption for many types of businesses simply because many businesses lack the resources and expertise to coordinate the development of a voice enabled application. Consequently, an important modality of interaction oftentimes is omitted from Web distributed content and customers suffer as a result.

BRIEF SUMMARY OF THE INVENTION

Embodiments of the present invention address deficiencies of the art in respect to the voice enablement of a Web page and provide a novel and non-obvious method and computer program product for the automated voice enablement of a Web page with free form input field support. In an embodiment of the invention, a method for voice enabling a Web page with free form input field support can be provided. The method can include receiving speech input for one or more input fields in a Web page, parsing a core attribute for the input field and identifying an external statistical language model (SLM) referenced by the core attribute of the input field, posting the received speech input and the SLM to an automatic speech recognition (ASR) engine, and inserting a textual equivalent to the speech input provided by the ASR engine in conjunction with the SLM into the input field.

In one aspect of the embodiment, the method further can include receiving additional speech input for a different input field in the Web page, parsing a core attribute for the different input field and identifying permitted terms set forth in the core attribute of the different input field. The method also can include generating a speech grammar for the different input field based upon the terms in the core attribute of the different input field and posting the additional speech input and the generated speech grammar to an ASR engine. Finally, the method can include inserting a textual equivalent to the additional speech input provided by the ASR engine in conjunction with the speech grammar into the different input field.

Additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The aspects of the invention will be realized and attained by means of the elements and combinations particularly pointed out in the appended claims. It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention, as claimed.

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS

The accompanying drawings, which are incorporated in and constitute part of this specification, illustrate embodiments of the invention and together with the description, serve to explain the principles of the invention. The embodiments illustrated herein are presently preferred, it being understood, however, that the invention is not limited to the precise arrangements and instrumentalities shown, wherein:

FIG. 1 is a pictorial illustration of a process for the automated voice enablement of a Web page;

FIG. 2 is a schematic illustration of a Web application data processing system configured for the automated voice enablement of a Web page; and,

FIG. 3 is a flow chart illustrating a process for voice enabling a Web page in a Web application data processing system with free form field support.

DETAILED DESCRIPTION OF THE INVENTION

Embodiments of the present invention provide a method and computer program product for the automated voice enablement of a Web page with free form input field support. In accordance with an embodiment of the present invention, a Web page of a Web application can be rendered in a content browser and voice input can be received for one or more input fields of the Web page. A grammar entry for a grammar can be determined for the input field, and if the grammar entry cannot be determined, a grammar entry can be dynamically generated for the input field. In this regard, a core attribute of the input field, such as the title of the input field, can specify permissible input for the input field, or the core attribute can reference an existing grammar.

In either case, the grammar and the voice input can be posted to a speech recognition engine from the content browser and a textual equivalent can be received in return. For instance, a REST compliant speech recognition engine can receive the posting of the voice input and grammar from the content browser and the speech recognition engine can return to the content browser a textual equivalent to the voice input. Thereafter, the textual equivalent can be inserted into a document object model (DOM) for the Web page and an affected portion of the Web page can be re-rendered to provide a visualization of the text input for the input field.

In this way, the Web application can be voice enabled even in the absence of specific expertise in speech application development. In particular, one seeking to voice enable fields of a Web page in a Web application need only specify permitted terms for the input field within a core attribute of the input field. The speech grammar entry can be generated automatically with reference to the core attribute of the input field. Further, the textual equivalent of voice input for the input field can be seamless integrated into the input field by inserting the textual equivalent directly into the DOM for the Web page. Thus, no speech recognition expertise is required to speech enable the Web page of the Web application.

Notably, the voice enablement of the input fields of the Web page can process not only the closed grammar of input fields of limited permitted language, but also the substantial grammar of a free form input field. In this regard, the core attribute of the input field can specify not only permissible input for the input field, or the core attribute can reference an existing grammar, but also the core attribute can reference an external statistical language model (SLM) suitable for handling speech input for a free form input field. In this way, the Web page can accommodate input fields of a limited grammar and those that are free form in nature.

In illustration, FIG. 1 pictorially shows a process for the automated voice enablement of a Web page. As shown in FIG. 1, a Web page 120 of a Web application (not shown) can be represented by DOM 130 produced by a content browser (not shown) rendering the Web page 120. The DOM 130 can define within the Web page 120 one or more different input fields 180A, 180B, 180N. Each of the input fields 180A, 180B, 180N can include a core attribute, such as a title. It is to be recognized by the skilled artisan that core attributes of markup language element like an input field are well-known and well-documented according to the Hypertext Markup Language (HTML) specification.

Voice enablement widget 190 can process speech input 150 provided by end user 110 for a given one of the input fields 180A, 180B, 180N. Specifically, upon received of speech input 150 from end user 110, it can be determined whether or not a grammar has been specified for the given one of the input fields 180A, 180B, 180N. If not, a grammar 140 can be generated for the input fields 180A, 180B, 180N by referring to a corresponding core attribute for each of the input fields 180A, 180B, 180N. In this regard, an exemplary markup language element for the input fields 180A, 180B, 180N providing permitted terms for a corresponding grammar follows:

<form dojoType=“VoiceForm” widgetID=“command” method=“post”>
Field_1: <input type=“text” name=“Field 1” title=“term1, term2,
term3 ”/>
Field_2: <input type=“text” name=“Field 2” title=“term4, term5,
term6 ”/>
Field_N: <input type=“text” name=“Field N” title=“term7, term8”/>
</form>

Additionally, a prefix can be provided for each field such as [the] or [an] or [to] to permit natural speech. Yet further, semantic meaning can be provided for the permitted terms. As an example,

<form dojoType=“VoiceForm” widgetID=“command” method=“post”>
Field_1: <input type=“text” name=“Field 1” title=“[the] term1, term2, term3 ”/>
Field_2: <input type=“text” name=“Field 2” title=“term4, term5, term6 ”/>
Field_N: <input type=“text” name=“Field N” title=“(to|$NULL),term7=1, term8=0”/>
</form>
\

In this way, Semantic Interpretation for Speech Recognition (SISR) can be provided for addition to the generated grammar so that the spoken form of a term can be translated to a value within a textual equivalent.

The resulting grammar 140 can reflect the permitted terms, prefix and semantic values, for example as follows:

#ABNF 1.0;
language en-US
mode voice;
tag-format <semantics/1.0>;
root $command;
met “author” is “VoiceGrammar Widget”;
public $command=$prologue $action ($Field_1 $Field_2 $Field_N)
$epilogue {$.Field_1=$Field_1; $.Field_2=$Field_2; $.Field_N=$Field_N}
$prologue=please | would you | $NULL
$action = do1 | do2 | doN | $NULL
$Field_1 = ([the] term1 {$=“term_1”} | term2 {$=“term_2} | term3 {$=“term_3}”/>
$Field_2 = (term4 {$=“term_4”} | term2 {$=“term_5} | term3 {$=“term_6}”/>
$Field_N = ((to|$NULL) term7 {$=“term_7”} | term8 {$=“term_8}”/>
$epilogue=thanks | now | $NULL;

Optionally, the resulting grammar 140 can be generated only for those of the input fields 180A, 180B, 180N that lack input provided by the end user 110.

Free form input fields also can be supported in automated voice enablement. Specifically, a given one of the input fields 180A, 180B, 180N can include a corresponding core attribute referencing an external SLM configured to support free form voice input. In this regard, an exemplary markup language element for the input fields 180A, 180B, 180N supporting free form input follows:

<form dojoType=“VoiceForm” widgetID=“command” method=“post”>
Field_1: <input type=“text” name=“Field 1” title=“term1, term2, term3 ”/>
Field_2: <input type=“text” name=“Field 2” title=“term4, term5, term6 ”/>
Field_N: <input type=“text” name=“Field N” title=“http://< domain.com/xyz.slm”/>
resulting in a grammar as follows:
#ABNF 1.0;
language en-US
mode voice;
tag-format <semantics/1.0>;
root $command;
met “author” is “VoiceGrammar Widget”;
public $command=$prologue $action ($Field_1 $Field_2 $Field_N)
$epilogue {$.Field_1=$Field_1; $.Field_2=$Field_2; $.Field_N=$Field_N}
$prologue=please | would you | $NULL
$action = do1 | do2 | doN | $NULL
$Field_1 = (term1 {$=“term_1”} | term2 {$=“term_2} | term3 ($=“term_3}”/>
$Field_2 = (term4 {$=“term_4”} | term2 {$=“term_5} | term3 ($=“term_6}”/>
$Field_N = ((to|$NULL) $<http://domain.com/xyz.slm>;/>
$epilogue=thanks | now | $NULL;

The voice enablement widget 190 can provide the resulting grammar 140 with the speech input 150 to a speech recognition service 160, for example a RESTful automatic speech recognition (ASR) engine. The speech recognition service 160 can process the speech input 150 in connection with the grammar 140 to produce a textual equivalent 170 to the speech input 150. The textual equivalent 170 can be returned to the voice enablement widget 190 which in turn can insert the textual equivalent 170 into a portion of the DOM 130 for the Web page 120 corresponding to the given one of the input fields 180A, 180B, 180N. Finally, the DOM 130 can be processed by the content browser (not shown) to re-render an affected portion of the Web page 120 with the textual equivalent 170 shown in the given one of the input fields 180A, 180B, 180N.

The process described in respect to FIG. 1 can be implemented in a Web application data processing system. In illustration, FIG. 2 schematically depicts a Web application data processing system configured for the automated voice enablement of a Web page. The system can include a content server 210 communicatively coupled to a client 240 over computer communications network 230. The content server 210 can be configured to serve Web pages 220 to the client 240 over the computer communications network 230 for rendering in a content browser 250. Optionally, the Web pages 220 each can be a “mash up” 260 of content drawn from multiple different content sources (not shown) distributed about the computer communications network 230.

Voice enablement widget 300 can be coupled to the content browser 250. The voice enablement widget 300 can include computer program code enabled to generate a grammar for providing voice input 270 for a field amongst input fields in the mash up 260 from terms in a core attribute of the input fields. The program code further can be enabled to provide both the grammar and the voice input 270 to an ASR engine 290 and to receive in return the text input 280 that is equivalent to the voice input 270. Finally, the program code can be yet further enabled to insert the text input 280 into a DOM for the mash up 260 and to re-render the affected portion of the mash up 260 in the content browser 250 to include the text input 260 in the input field of the mash up 260 into which the voice input 270 had been provided.

In yet further illustration of the operation of the voice enablement widget 300, FIG. 3 is a flow chart illustrating a process for voice enabling a Web page in a Web application data processing system with free form input field support. Beginning in block 305, a Web page can be received for rendering in a content browser and in block 310 a DOM can be created for the Web page. Subsequently, in block 315 the DOM can be rendered in the content browser as a Web page.

The Web page can provide one or more input fields into which textual input is expected, however, a user interface element can be provided in the Web page in connection with one or more of the input fields that when activated permits speech input to be provided in place of textual input for the input field. As such, in block 320, an input field can be selected for speech input and in decision block 325 it can be determined whether or not a speech grammar exists for the input fields. If so, the grammar can be loaded. Otherwise, in block 330 terms a core attribute of each of the input fields can be retrieved and in decision block 370, it can be determined whether or not the input field is free form. If not, in block 335 a speech grammar can be generated based upon terms set forth by the core attributes for the input fields. However, if it is determined that the input field is free form, in block 380 an SLM grammar reference by the core attribute can be retrieved as the grammar for the input field.

In any event, in block 345 speech input can be received for the input field and in block 350, the speech input can be posted to a coupled ASR engine along with the speech grammar. Thereafter, in block 355 equivalent text produced by the ASR engine for the speech input can be inserted into a node for the DOM corresponding to the input field. Finally, in block 360 the Web page can be re-rendered to show the textual input in the input field. Notably, because main driving grammar of the Web page is an ordinary grammar and contains semantic interpretation, if an input field is resolved via an external SLM, the semantic interpretation still can be applied to the speech recognized text despite the use of an SLM.

Embodiments of the invention can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment containing both hardware and software elements. In a preferred embodiment, the invention is implemented in software, which includes but is not limited to firmware, resident software, microcode, and the like. Furthermore, the invention can take the form of a computer program product accessible from a computer-usable or computer-readable medium providing program code for use by or in connection with a computer or any instruction execution system.

For the purposes of this description, a computer-usable or computer readable medium can be any apparatus that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. The medium can be an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system (or apparatus or device) or a propagation medium. Examples of a computer-readable medium include a semiconductor or solid state memory, magnetic tape, a removable computer diskette, a random access memory (RAM), a read-only memory (ROM), a rigid magnetic disk and an optical disk. Current examples of optical disks include compact disk-read only memory (CD-ROM), compact disk-read/write (CD-R/W) and DVD.

A data processing system suitable for storing and/or executing program code will include at least one processor coupled directly or indirectly to memory elements through a system bus. The memory elements can include local memory employed during actual execution of the program code, bulk storage, and cache memories which provide temporary storage of at least some program code in order to reduce the number of times code must be retrieved from bulk storage during execution. Input/output or I/O devices (including but not limited to keyboards, displays, pointing devices, etc.) can be coupled to the system either directly or through intervening I/O controllers. Network adapters may also be coupled to the system to enable the data processing system to become coupled to other data processing systems or remote printers or storage devices through intervening private or public networks. Modems, cable modem and Ethernet cards are just a few of the currently available types of network adapters.